Radiation of bilateral adrenal metastases is associated with a high risk of primary adrenal insufficiency

Summary:

Background:

Adrenal metastasis is the most common adrenal malignancy and can be bilateral in up to 43% of patients. Radiotherapy (RT) is one option available to treat adrenal metastases. The risk of PAI after adrenal RT is unclear.

Objective:

Determine the incidence and the timeline of PAI in patients undergoing adrenal RT.

Design, Setting, and Participants:

Single-center longitudinal retrospective cohort study of adult patients with adrenal metastases treated with RT between 2010 and 2021.

Results:

Of 56 patients with adrenal metastases treated with adrenal RT, 8 (14.3%) patients developed PAI at a median of 6.1 months (IQR 3.9–13.8) after RT All patients developing PAI had either unilateral RT in the setting of contralateral adrenalectomy or bilateral adrenal RT. Patients who developed PAI received a median RT dose of 50 Gy (IQR 44–50 Gy), administered in a median of 5 fractions (IQR 5–6). Treated metastases decreased in size and/or metabolic activity on positron emission tomography in 7 patients (87.5%). Patients were initiated on hydrocortisone (median daily dose of 20 mg, IQR 18–40) and fludrocortisone (median daily dose of 0.05 mg, IQR 0.05–0.05 mg). At the end of the study period, five patients died, all due to extra-adrenal malignancy, at a median time of 19.7 months (IQR 16–21.1 months) since RT and median time of 7.7 months (IQR 2.9–12.5 months) since the diagnosis of PAI.

Conclusion:

Patients receiving unilateral adrenal RT with two intact adrenal glands have a low risk of PAI. Patients receiving bilateral adrenal RT have a high risk of PAI and require close monitoring.

Introduction

Adrenal metastasis is the most common adrenal malignancy encountered in clinical practice, accounting for approximately 7.5% of all adrenal masses, and 85–90% of malignant adrenal masses^1,2. While many are diagnosed during cancer staging imaging, approximately one third are found incidentally^1,3. Bilateral disease is common, including approximately 24% of patients at initial diagnosis and up to up to 43% will ultimately develop bilateral adrenal metastases³. Primary adrenal insufficiency (PAI) is diagnosed in 12.7% of all patients with bilateral adrenal metastases, and in 20% if adrenal metastases are > 4 cm³. Adrenal radiotherapy (RT) is a standard therapeutic modality employed in the treatment of adrenal metastases and is generally effective and well-tolerated^4–15. In addition to the already high risk of PAI in patients with bilateral adrenal metastases, superimposed adrenal RT may contribute to the residual adrenal cortex destruction through direct cellular and DNA damage¹⁶.

Few studies have described the incidence of PAI in patients with adrenal RT^{5,6,10,11,13,15,17,18} though relationship to RT was not always conclusive due to concomitant exogenous glucocorticoid use or other systemic treatments that may induce PAI⁵. Available data are further limited by incomplete characterization of the time to PAI onset after RT, clinical presentation, and details on PAI diagnosis and management.

The aim of our study was to determine the incidence, and the timeline for development of PAI in patients undergoing adrenal RT.

Methods

Patient Selection

The study protocol was approved by our Institutional Review Board. We used an institutional medical record search software on January 3^rd, 2022, to identify eligible patients using the following terms in various combinations: “Adrenal Metastasis”, “Bilateral Adrenal Metastases”, “Primary Adrenal Insufficiency” and “Radiation”. Patients were included in the study for case review if they: 1) were evaluated at our institution between January 1^st, 2010, and December 31^st, 2021; 2) had a history of adrenal metastases suspected based upon the positron emission tomography (PET) results, increase in size of adrenal mass based on cross-sectional abdominal imaging, and/or confirmed adrenal metastasis upon biopsy, 3) had undergone unilateral or bilateral RT of adrenal metastasis/es, and 4) did not have PAI at the time of adrenal RT. Patients were divided into two groups; Group 1 was RT to all present adrenal glands (bilateral RT or RT to a solitary adrenal gland) and Group 2 was unilateral RT in the setting of present bilateral adrenal glands, Figure 1.

Figure 1.

Selection of Patients for Study Eligibility

^aPatients with bilateral adrenal radiation or unilateral radiation of solitary adrenal gland

^bPatients with unilateral adrenal radiation in the setting of present bilateral adrenal glands

Data Collection

Information on primary cancer, tumor size, imaging characteristics, RT dose and fractionation, biochemical parameters, date of PAI diagnosis and medical management was collected. Biochemical parameters included plasma sodium (normal 135–145 mEq/L), potassium (normal 3.6–5.2 mEq/L), morning cortisol (normal 7–25 mcg/dL), adrenocorticotropic hormone ([ACTH] normal 7.2–63 pg/mL), renin plasma activity (normal 0.6–4.0 ng/mL/hr, sodium replete, upright) and aldosterone (normal < 21 ng/dL). Imaging was reviewed at the time of RT, and during follow up. PAI was diagnosed based on: 1) biochemical criteria of elevated ACTH and low morning cortisol for glucocorticoid deficiency; 2) hyponatremia and/or hyperkalemia with our without elevated renin level for mineralocorticoid deficiency; 3) and/or clinical evaluation.

Statistical Analysis

Categorical data were presented as counts and percentages and continuous data were presented as median and interquartile ranges.

Results

Eligible Patients

A total of 56 patients met eligibility criteria and were included for analysis, Table 1. Ten patients were in Group 1 (9 patients had bilateral RT, 1 patient had solitary adrenal gland RT), and 46 patients were in Group 2. Diagnosis of PAI was made based on biochemical parameters and/or clinician decision. Final cohort of patients (n = 8, 14.3%), were all from Group 1, Figure 1. One of the patients in Group 1 who did not develop PAI passed away 6 months after radiation therapy, the other is alive 24 months post-radiation. No differences in age, sex, RT dose and fractions were found between patients in Group 1 and 2 (data not shown). Group 2 was not analyzed further in this study as no PAI was reported in this group.

Table 1.

Characteristics of patients undergoing adrenal radiation

Age, years (Median, [IQR])	69.5 (64–79.3)
Sex (n, %)
Female	22 (39.3%)
Male	34 (60.7%)
Primary Tumor (n, %)
Breast	1 (1.8%)
Colorectal	1 (1.8%)
GE Junction	1 (1.8%)
Lung	41 (73.2%)
Lymphoma	1 (1.8%)
Melanoma	4 (7.1%)
Neuroendocrine	1 (1.8%)
Ovarian	1 (1.8%)
Pancreas	1 (1.8%)
Prostate	1 (1.8%)
Renal Cell	3 (5.3%)
Radiation Laterality (n, %)
Bilateral1	9 (16.1%)
Unilateral, solitary adrenal1	1 (1.8%)
Unilateral2	46 (82.1%)
Radiation Dose, Gy (Median, [IQR])	50 (40–50)
Radiation Fractions, # (Median, [IQR])	5 (5–5)

¹Group 1 for the study

²Group 2 for the study

Abbreviations: GE (Gastroesophageal), Gy (Gray), IQR (Interquartile Range), # (number)

Characteristics of Patients with Primary Adrenal Insufficiency

The median age at the time of adrenal RT was 69 years (IQR 62–72 years) with median follow-up of 17.9 months (IQR 8.1–29.8 months). Sex and primary malignancy are noted in Table 2. Adrenal metastases were diagnosed based on biopsy in four patients (50%), and the others were diagnosed based on increase in size and/or increase in uptake within the adrenal glands during cancer staging imaging. At the time of RT, median size of left adrenal metastases (n = 7) was 25 mm (IQR 13.75–25 mm), and median size of the right adrenal metastases (n = 6) was 19 mm (IQR 3.75–25 mm). Median RT dose administered was 50 Gy (IQR 44–50 Gy), an ablative dose, in a median of five fractions (IQR 5–6), Table 2. RT was noted to be effective in 7 patients (87.5%), as treated metastases either decreased in size or demonstrated a decrease in metabolic activity during median imaging follow up, Table 2. All patients had been referred to endocrinology for monitoring/assessment of adrenal insufficiency. Of those patients, five (62.5%) were referred prior to radiation. Follow-up assessment frequency varied by provider, ranging between every 3–12-months. By the end of the study period, five patients (62.5%) had died, a median time of 19.7 months post-RT (IQR 16–21.1 months), Table 2.

Table 2.

Clinical, biochemical, and therapy data of patients diagnosed with primary adrenal insufficiency

	Sex	Age	Primary Malignancy	Adrenal metastasis size at time of radiation (mm)	Radiation Dose (Gy)	Number of Radiation Fractions	Adrenal metastasis at the time of last follow-up (mm)	Time Since Radiation to Diagnosis of PAI (months)	Symptoms at the Time of Diagnosis	Initial daily dose of glucocorticoid and mineralocorticoid replacement (mg)	Last documented daily dose of glucocorticoid and mineralocorticoid replacement (mg)	Time since radiation to last follow-up (months)
Biochemical Diagnosis
Patient 1	F	72	Lung	L: 28 R: 34	50	5	L: 18 R: 14	20.4	Yes	HC: 15 FC: 0.05	HC: 15 FC: 0.075	60.8
Patient 2	M	72	Lung	L: 22 R: 26	20	5	L: 14 R: 49	7.2	Yes	HC: 20 FC: 0.05	HC: 20 FC: 0.05	19.7a
Patient 3	F	63	Lung	L: 15 R: 02	50	5	L: 8 R: 03	14.8	Yes	HC5: x FC5: x	HC5: x FC5: x	16a
Patient 4	M	69	Lung	L: 10 R: 19	50	5	L: 9 R: 10	4.1	Yes	HC6: 40 FC: 0.05	HC6: 40 FC: 0.1	4.4
Patient 51	M	68	Melanoma	L: 40 R: 21	40	5	L: 44 R: 20	4.9	No	HC: 15 FC: x	HC: 20 FC: x	7.9
Clinical Diagnosis
Patient 6	F	59	Lung	L: 25 R: 17	45	18	L: x4 R: 11	0.5	No	HC: 20 FC: 0.1	HC: 35 FC: 0.1	55.8a
Patient 7	M	58	Lung	L: 02 R: 02	65	Unknown	L: 03 R: 03	3.2	Yes	HC: 80 FC: 0.05	HC: 160 FC: 0.05	6.1a
Patient 8	F	78	Lung	L: 25 R: 25	50	5	L: 17 R: 7	13.4	Yes	HC: 40 FC: x	HC: 80 FC: x	21.1a

Abbreviations: F (Female), FC (fludrocortisone), Gy (Gray), HC (hydrocortisone), L (left), M (male), R (right)

^aPatient expired at the time of study conclusion

¹Previous radiation to left adrenal, completed right adrenal radiation at the time of study enrollment

²Increased uptake in the adrenal gland, no nodule noted, metastasis presumed

³Decreased uptake in adrenal gland

⁴Adrenalectomy

⁵Patient had was diagnosed with partial adrenal insufficiency; utilized sick day rules for HC only, no daily dose

⁶Patient was on prednisone 10 mg for a separate side effect related to immune checkpoint inhibitor therapy

Primary Adrenal Insufficiency Diagnosis

Median time to diagnosis of PAI after RT was 6.1 months (IQR 3.9–13.8 months). Of the eight patients who developed PAI, seven (87.5%) had been followed by endocrinology at the time of RT and PAI was anticipated. Two patients (25%) were diagnosed based on symptoms consistent with PAI given the clinical context of adrenal metastases and RT, and one patient (12.5%) was pre-emptively treated after RT was completed due to history of previous unilateral adrenalectomy. Five patients (62.5%) were diagnosed based on biochemical work-up consistent with PAI, of whom four patients also had symptoms of PAI. Three of those patients (60%) were diagnosed on planned biochemical monitoring, and two patients (40%) were diagnosed serendipitously during a hospitalization. One patient was diagnosed when hospitalized for nausea, vomiting, and diarrhea, demonstrating a cortisol level of 7.8 mcg/dL given the patient’s acute presentation. Another patient was hospitalized for back pain and was found to be hyponatremic and hyperkalemic (Table 3). Seven patients (87.5%) were treated with immune checkpoint inhibitors and/or glucocorticoids at some time prior to the diagnosis of PAI without prior diagnosis of central adrenal insufficiency.

Table 3.

Biochemical Assessment

Test Description (Units)	Baseline Lab Tests Prior to Radiation	Lab Tests at the time of PAI Diagnosis	Reference Range
Sodium (mEq/L)			135–145
Median (IQR)	141 (139–143)	134 (133–137)
Patient 1	139	132
Patient 2	144	139
Patient 3	144	134
Patient 4	139	132
Patient 5	140	139
Patient 6	143	-
Patient 7	141	135
Patient 8	141	134
Potassium (mEq/L)			3.8–5.2
Median (IQR)	4.7 (4.4–4.8)	5.0 (4.5–5.5)
Patient 1	4.9	5.6
Patient 2	4.5	4.8
Patient 3	4.8	3.9
Patient 4	4.3	5.8
Patient 5	4.4	4.2
Patient 6	1.8	-
Patient 7	4.8	5.0
Patient 8	4.8	5.4
Adrenocorticotropic Hormone (pg/mL)			7.2–63
Median (IQR)	21 (20–26)	124 (70–311)
Patient 1	-	498
Patient 2	41	124
Patient 3	21	-
Patient 4	-	-
Patient 5	-	-
Patient 6	17	-
Patient 7	-	-
Patient 8	-	151
Morning cortisol (mcg/dL)			7–25
Median (IQR)	12.3 (10.6–14.7)	6.4 (6.1–7.8)
Patient 1	-	6.1
Patient 2	15	9.9
Patient 3	10	7.8
Patient 4	-	-
Patient 5	10.6	6.4
Patient 6	19	-
Patient 7	-	-
Patient 8	-	1.0
Renin (ng/mL/hr)			0.6–4.0 (sodium
Median (IQR)	2.2 (1.7–2.6)	8.0 (2.5–13.5)	replete, upright)
Patient 1	-	13
Patient 2	-	1.3
Patient 3	1.1	-
Patient 4	2.2	2.9
Patient 5	3.0	-
Patient 6	-	-
Patient 7	-	15
Patient 8	-	-
Aldosterone (ng/dL)			< 21
Median (IQR)		4 (4–7.5)
Patient 1		4
Patient 2		-
Patient 3		-
Patient 4		4
Patient 5		-
Patient 6		-
Patient 7		11
Patient 8		-

¹Patient was on prednisone, potentially impacting result

Abbreviations: IQR (Interquartile Range); PAI (Primary Adrenal Insufficiency)

Prior to the diagnosis of PAI, median cortisol (n = 5) was 12.3 mcg/dL (IQR 10.6–14.7 mcg/dL), median ACTH (n = 4) was 21 pg/mL (IQR 20–26 pg/mL), median sodium (n = 8) was 141 mEq/L (IQR 139–143 mEq/L), median potassium (n = 8) was 4.7 mEq/L (IQR 4.4–4.8 mEq/L) and median renin plasma activity (n = 3) was 2.2 ng/mL/hr (IQR 1.7–2.6 ng/mL/hr). At the time of PAI diagnosis, median cortisol (n = 5) was 6.4 mcg/dL (IQR 6.1–7.8 mcg/dL), median ACTH (n = 3) was 124 pg/mL (IQR 70–311 pg/mL), median sodium (n = 7) was 134 mEq/L (IQR 133–137 mEq/L), median potassium (n = 7) was 5.0 mEq/L (IQR 4.5–5.5 mEq/L), median renin plasma activity (n = 4) was 8.0 ng/mL/hr (IQR 2.5–13.5 ng/mL/hr), and median aldosterone (n = 3) was 4 ng/dL (IQR 4–7.5 ng/dL), Table 3. Patient 1 also completed a cosyntropin stimulation test with suboptimal response; with a baseline cortisol of 7.9 mcg/dL and a 60-minute stimulated cortisol of 13 mcg/dL. Patient 6 did not have any diagnostic work-up completed as they were empirically treated with adrenal replacement therapy after their radiation treatment was completed.

Primary Adrenal Insufficiency Treatment

One patient (patient 3) was not treated with daily hydrocortisone or fludrocortisone as PAI was considered mild, however received education on sick day rules as well as symptoms of PAI necessitating daily therapy. Glucocorticoid replacement was either hydrocortisone or prednisone, depending on treating provider’s preference. Initial total daily glucocorticoid dose (n = 7, 87.5%), hydrocortisone equivalent, was 20 mg (IQR 18–40 mg). Initial total daily fludrocortisone dose (n = 5, 62.5%) was 0.05 mg (IQR 0.05–0.05 mg). One patient had fludrocortisone started four months after glucocorticoids due to symptoms likely related to mineralocorticoid deficiency and prior diagnosis of glucocorticoid deficiency. Last known total daily glucocorticoid dose at the end of follow-up (n = 7, 87.5%), hydrocortisone equivalent, was 35 mg (IQR 20–60 mg) and last known total daily fludrocortisone dose (n = 6, 75%) was 0.05 mg (IQR 0.05–0.075 mg), Table 1. Four patients were on supraphysiologic glucocorticoid therapy at the end of follow-up due to treatment of immune-checkpoint inhibitor side effect (n = 1, 12.5%), treatment of chronic obstructive pulmonary disease (n = 1, 12.5%), and persistent symptoms considered to be related to PAI with dose increased by clinician (n = 1, 12.5%) or self-increased by the patient (n = 1, 12.5%).

Discussion

This study provides detailed information about the risk of PAI in patients undergoing adrenal RT with specific information regarding clinical and laboratory presentation, time of onset, and treatment of glucocorticoid and mineralocorticoid deficiency.

Adrenal RT was successful in treating the adrenal metastases in our cohort, as most patients had a decrease in the size(s) of their adrenal metastatic lesions on subsequent imaging and/or decrease in the metabolic activity on the PET scans. However, we found a significantly higher rate of primary adrenal insufficiency compared to what has been reported in the literature, with 80% of those undergoing RT of all present adrenal glands, generally within one-year post-RT. Reporting of PAI was generally limited in prior studies, with a lower incidence of 50% or less (if reported), limited indication of timeframe to development of PAI or how PAI was diagnosed, and relationship to RT was not always conclusive due to concomitant exogenous glucocorticoid use or other systemic treatments that may induce adrenal insufficiency^{4–13,15,17,19}.

Time to development of PAI was relatively short at a median of 6.1 months in our cohort. Most patients in our study were followed by an endocrinology provider in anticipation of developing PAI at some point after their RT was completed, although it was still noted that a few patients were diagnosed at a time that was unexpected, such as during a hospitalization for other reasons, and rarely on biochemical monitoring alone prior to significant symptom development. Compared to RT of most other endocrine organs, PAI appears to develop sooner after treatment with RT, and at higher incidence, if specifically referring to RT of all present adrenal glands. RT delivered to the pituitary gland or near the pituitary gland (e.g., for nasopharyngeal carcinomas, other skull based tumors) resulting in hypopituitarism has been reported to be more of a “late effect” (median time to hypopituitarism not always reported) with incidence ranging between 24.2%−66% of patients after 5–10 years^20–24, although some report nearly 100% incidence²². Hypothyroidism has been reported at a median of 15.4–46 months post RT of the neck, with incidence of 38.0–54.7%^25,26. Incidence of diabetes mellitus after abdominal RT and male hypogonadism after pelvic radiation is low^27–29; in contrast, 100% of premenopausal females who undergo pelvic RT without prior ovarian transposition out of the RT field will develop primary ovarian insufficiency³⁰. It is possible that PAI developed sooner and at higher incidence in patients with RT to all present adrenal glands compared to most other endocrine organs as there may have been destruction of the adrenal cortex because of the metastatic disease. Additionally, data has suggested a correlation between RT fraction size and total dose with the risk of endocrine dysfunction, or on the contrary, endocrine normalization in the context of hypersecretion. Extrapolating from available data from analogous disease sites like pituitary adenomas, the median dose of 50 Gy in 5 fractions utilized in this study would be associated with a relatively rapid rate of endocrine dysfunction^31,32.

Irrespective of adrenal RT status, all patients with bilateral adrenal infiltrative disease, and especially those with signs and symptoms consistent with PAI should be evaluated with morning cortisol and ACTH^1,33. In our study, once patients developed PAI from RT, treatment with glucocorticoid and mineralocorticoid replacement therapy was initiated in most patients as the first step; it was less common for mineralocorticoid deficiency to present later. A few patients were on supraphysiologic glucocorticoid therapy for reasons not related to PAI, and several patients had their dose increased in attempt to treat symptoms considered to be related to PAI. Thus, ongoing endocrine follow up and collaboration with oncology team, as well as reenforcement of sick day rule education is beneficial. We have developed a protocol for patients undergoing RT of all present adrenal glands regarding baseline education and testing, follow-up monitoring, and after diagnosis management in Table 4.

Table 4.

Recommended Follow-Up for Adrenal Radiation Patients

Baseline Patient education regarding risk and symptoms of PAI Baseline electrolytes (sodium, potassium), ACTH, 8 AM cortisol, and renin plasma activity

Every 3–6 Months Clinical assessment for signs and symptoms of PAI Laboratory testing including electrolytes, ACTH, 8 AM cortisol, renin plasma activity, and aldosterone

After Diagnosis Initiation of physiologic glucocorticoid replacement Initiation of mineralocorticoid replacement when applicable Patient education regarding stress-dosing/sick day rules for glucocorticoid replacement Establish follow-up plan with clinical assessment of symptom control and laboratory testing to include electrolytes +/− renin plasma activity

Abbreviations: ACTH (adrenocorticotropic hormone), PAI (primary adrenal insufficiency)

Our study has several strengths, such as detailed characterization of clinical, biochemical and management of patients undergoing bilateral adrenal RT. Limitations include heterogeneity in diagnosis and monitoring of patients, and those related to the retrospective design of the study, such as information bias and selection bias.

Conclusion

Adrenal RT is an effective treatment for adrenal metastasis. In patients treated with bilateral adrenal RT or unilateral adrenal RT in the setting of single adrenal gland, the risk of PAI is 80% with the median time of onset at 6 months post-RT. These patients should receive counseling about the risk and symptoms of PAI, education on PAI management, and establish a close endocrine follow up and monitoring for PAI. Patients receiving unilateral adrenal RT with two intact adrenal glands have a low risk of PAI.

Data Availability Statement:

Some or all data sets generated during and/or analyzed during the present study are not publicly available but are available from the corresponding author on reasonable request.